Stroke (CVA) is the leading cause of disability in the United States and it is estimated that its prevalence will more than double over the next 50 years. Current stroke therapy is labor-intensive and costly. The United States spends $17 billion taking care of stroke survivors. Recently, concentrated, massed practice therapies have been developed that improve function in CVA patients by reversing the effects of "learned nonuse." The objective of this project is to investigate the feasibility of a device that facilitates the administration of massed practice stroke therapy. The long-term objective is to provide a lightweight device for home use that provides motion and biofeedback of desired and undesirable muscle activity. Software controls the function of the device and monitors patient progress and compliance. A pneumatic artificial muscle will be used to provide physical motion. This artificial muscle has many of the properties of human muscle. It is lightweight, flexible and has spring like properties. This project will focuses on treating wrist and finger extensor weakness, however, the concept applies to all areas affected by motor impairment. This Phase I study includes detailed design verification measurements on the device and measures the responses of able bodied test subjects to the treatment protocol.